Preceramic Inorganic Polymers

One of the most important interfaces in materials science is the one between polymers and ceramics. Ceramics can be viewed as highly cross-linked polymer systems, with the three-dimensional network providing strength, rigidity, and resistance to high temperatures. Although not generally recognized as such, a few ceramics exist that are totally organic (i.e., carbon-based). Melamine-formaldehyde resins, phenolformaldehyde materials, and carbon fibers are well-known examples. However, totally inorganic ceramics are more widely known, many of which are based on the elements silicon, aluminum, or boron combined with oxygen, carbon, or nitrogen. Among the inorganic ceramics, two different classes can be recognized—oxide ceramics and non-oxide materials. The oxide ceramics frequently include silicate structures, and these are relatively low melting materials. The non-oxide ceramics, such as silicon carbide, silicon nitride, aluminum nitride, and boron nitride are some of the highest melting substances known. Non-oxide ceramics are often so high melting that they are difficult to shape and fabricate by the melt- or powder-fusion techniques that are common for oxide materials. One major use for inorganic-organic polymers and oligomers is as sacrificial intermediates for pyrolytic conversion to ceramics. The logic is as follows. Linear, branched, or cyclolinear polymers or oligomers can be fabricated easily by solution- or melt-fabrication techniques. If a polymeric material that has been shaped and fabricated in this way is then cross-linked and pyrolyzed in an inert atmosphere to drive off the organic components (typically, the side groups), the resultant residue may be a totally inorganic ceramic in the shape of the original fabricated article. Thus, ceramic fibers, films, coatings, and shaped objects may by accessible without recourse to the ultra-high temperatures needed for melting of the ceramic material itself. Note, however, that although the final shape of the object may be retained during pyrolysis, the size will be diminished due to the loss of volatile material. If the pyrolysis takes place too quickly, this contraction process may cause cracking of the material and loss of strength.

Polysiloxanes and Related Polymers

At the present time, polysiloxanes are unique among inorganic and semi-inorganic polymers. They have been the most studied by far, and are the most important with regard to commercial applications. Thus, it is not surprising that a large number of review articles exist describing the synthesis, properties, and applications of these materials. The Si-O backbone of this class of polymers endows it with a variety of intriguing properties. For example, the strength of this bond gives the siloxane polymers considerable thermal stability, which is very important for their use in high-temperature application (for example as heat-transfer agents and high-performance elastomers). The nature of the bonding and the chemical characteristics of typical side groups give the chains a very low surface free energy and, therefore, highly unusual and desirable surface properties. Not surprising, polysiloxanes are much used, for example, as mold-release agents, for waterproofing garments, and as biomedical materials. Some unusual structural features of the chains give rise to physical properties that are also of considerable scientific interest. For example, the substituted Si atom and the unsubstituted O atom differ greatly in size, giving the chain a very irregular cross section. This influences the way the chains pack in the bulk, amorphous state, which, in turn, gives the chains very unusual equation-of-state properties (such as compressibilities). Also, the bond angles around the O atom are much larger than those around the Si, and this makes the planar all-trans form of the chain approximate a series of closed polygons. As a result, siloxane chains exhibit a number of interesting configurational characteristics. These structural features, and a number of properties and their associated applications, will be discussed in this chapter. The major categories of homopolymers and copolymers to be discussed are linear siloxane polymers [-SiRR'O-] (with various alkyl and aryl R,R' side groups), (ii) sesquisiloxane polymers possibly having a ladder structure, (iii) siloxane-silarylene polymers [-Si(CH3)2OSi(CH3)2(C6H4)m-] (where the skeletal phenylene units are either meta or para), (iv) silalkylene polymers [-Si(CH3)2(CH2)m-], and (v) random and block copolymers, and blends of some of the above. Topics of particular importance are the structure, flexibility, transition temperatures, permeability, and other physical properties.

Inorganic-Organic Hybrid Composites

A relatively new area that involves silicon-containing materials is the synthesis of “ultrastructure” materials, that is materials in which structure can be controlled at the level of around 100 Å. An example of such a synthesis is the “sol-gel” hydrolysis of alkoxysilanes (organosilicates) to give silica, SiO2. The reaction is complicated, involving polymerization and branching, but a typical overall reaction may be written . . . Si(OR)4 + 2H2O → SiO2 + 4ROH (1) . . . where the Si(OR)4 organometallic species is typically tetraethoxysilane (tetraethylorthosilicate) (TEOS, with R being C2H5). In this application, the precursor compound is hydrolyzed and then condensed to polymeric chains, the chains become more and more branched, and finally a continuous highly swollen gel is formed. It is first dried at moderately low temperatures to remove volatile species, and then is fired into a porous ceramic object. It can then be densified, if desired, and machined into a final ceramic part. Not surprisingly, the production of ceramics by this novel route has generated a great deal of interest. Its advantages, over the usual “heat-and-beat” (e.g., sintering) approach to ceramics, is (i) the higher purity of the starting materials, (ii) the relatively low temperatures required, (iii) the possibility of controlling the ultrastructure of the ceramic (to reduce the number of microscopic flaws that lead to brittleness), (iv) the ease with which ceramic coatings can be formed, and (v) the ease with which ceramic alloys can be prepared (for example, by hydrolyzing solutions of both silicates and titanates). This approach has been used to form ceramic-like phases in a wide variety of polymers. The one which has been studied the most in this regard is poly(dimethylsiloxane) (PDMS), the semi-inorganic polymer featured extensively in Chapter 4. This is due to PDMS being in the class of relatively weak elastomers most in need of reinforcement, and being capable of easily absorbing the precursor materials generally used in the sol-gel process. The same hydrolyses can be carried out within a polymeric matrix to generate particles of the ceramic material, typically with an average diameter of a few hundred angstroms. The polymer typically has end groups, such as hydroxyls, that can participate in the hydrolysis-condensation reactions.

Ferrocene-Based Polymers, and Additional Phosphorus- and Boron-Containing Polymers

Ferrocene is an inexpensive, stable molecule with an interesting and reversible electrochemistry. It is synthesized by the metal-hydrogen exchange reaction of cyclopentadiene with sodium followed by treatment of the resultant sodium cyclopentadienide anion with ferrous chloride. The high stability and electroactivity of the ferrocene molecule has prompted numerous attempts to incorporate it into polymer structures. So, too, has the inherent torsional freedom of the cyclopentadienyl groups around the iron atoms and their capacity to serve as swivel group sites. Polymerization attempts range from the addition reactions of vinylferrocene and its derivatives, to condensation reactions, ringopening polymerizations, and dendrimer assemblies. These will be considered in turn. Considerable effort in the 1970s by Pittman, George, Hayes, Korshak, and others was applied to exploring the addition polymerization of vinylferrocene to give organic polymers with pendent ferrocenyl side groups. This type of polymerization reaction has been attempted with the use of free radical, cationic, anionic, and Ziegler–Natta methods. For free radical polymerization reactions, the initiating radicals must be generated from azo-initiators because peroxides cause oxidation of the metal. In polymerizations of the type shown in reaction (2) the side group ferrocene units are the source of both the thermal stability of the product polymers and complications inherent in the free radical polymerization process. For example, electron donation from the iron atoms to a growing radical chain end can convert an active radical to an anion, which terminates the polymerization. The Fe+ center then rearranges to form a paramagnetic, ionically bound Fe(III) species. Ultimately this leads to extensive chain-transfer, limitation of the chain length, and formation of branched structures. This does not occur if the ferrocene unit is insulated from the vinyl group by a spacer unit, as in, because these monomers polymerize normally. For example, monomer gives polymers with Mn molecular weights as high as 250,000. However, the electron-transfer process outlined in reaction (2) has serious practical consequences in the free-radical polymerization of. First, directly or indirectly, it causes precipitation of the growing polymer chains until, at monomer to polymer conversions of 90% or more, all of the polymeric product is insoluble in most organic solvents.

Polysilanes and Related Polymers

In polysilane polymers, the polymer backbone is made up entirely of silicon atoms. Therefore these materials differ from other important inorganic polymers, the siloxanes and phosphazenes, in which the polymer chain is heteroatomic. Structurally, they are more closely related to homoatomic organic polymers such as the polyolefins. However, because the units in the main chain are all silicon atoms, the polysilanes exhibit quite unusual properties. The cumulated silicon-silicon bonds in the polymer chain allow extensive electron delocalization to take place, and this delocalization of the sigma electrons in the Si-Si bonds gives the polysilanes unique optical and electronic properties. Many of the potential technical uses, as well as the remarkable properties, of polysilanes result from this unusual mobility of the sigma electrons. The polysilanes can be regarded as one-dimensional analogs to elemental silicon, on which, of course, nearly all of modern electronics is based. The photophysical behavior of polysilanes is not approached by any other materials, save for the less stable and more costly polygermanes and polystannanes. The remarkable properties of polysilanes have led to intense interest, and to numerous proposed high-tech applications. But the great promise of polysilanes as materials has yet to be realized. Their only commercial use at present is as precursors to silicon carbide ceramics, an application which takes no advantage of their optical or electronic properties. Linear polysilane polymers, properly called poly(silylene)s, can be obtained as homopolymers or copolymers. Continuation of the polysilane chain consumes two of the four valences of each silicon atom; the other two are taken up by pendent groups, which may be the same or different. Copolymers, which contain two or more kinds of silicon atoms, can be made up from units. A typical example is the copolymer of Me2Si and PhMeSi units, poly(dimethylsilylene-co-phenylmethylsilylene), which bears the popular name “polysilastyrene.” The pendent groups are typically organic units and can include alkyl, aryl, substituted aryl, hydrogen, Me3Si, ferrocenyl, and so on. An unlimited number of different polymers are possible, and several hundred compositions have been described in the literature.

Introduction

A polymer is a very-long-chain macromolecule in which hundreds or thousands of atoms are linked together to form a one-dimensional array. The skeletal atoms usually bear side groups, often two in number, which can be as small as hydrogen, chlorine, or fluorine atoms or as large as aryl or long-chain alkyl units. Polymers are different from other molecules because the long-chain character allows the chains to become entangled in solution or in the solid state or, for specific macromolecular structures, to become lined up in regular arrays in the solid state. These molecular characteristics give rise to solid-state materials properties, such as strength, elasticity, fiber-forming qualities, or film-forming properties, that are not found for small molecule systems. The molecular weights of polymers are normally so high that, for all practical purposes, they are nonvolatile. These characteristics underlie the widespread use of polymers in all aspects of modern technology. Attempts to understand the relationship between the macromolecular structure and the unusual properties characterize much of the fundamental science in this field. Polymers are among the most complicated molecules known. They may contain thousands of atoms in the main chain, plus complex clusters of atoms that form the side groups attached to the skeletal units. How, then, can we depict such molecules in a manner that is easy to comprehend? First, an enormous simplification can be achieved if we remember that most synthetic polymers contain a fairly simple structure that repeats over and over down the chain. This simplest repetitive structure is known as the repeating unit, and it provides the basis for an uncomplicated representation of the structure of the whole polymer. For example, suppose that a polymer consists of a long chain of atoms of type A, to which are attached side groups, R. The polymer chain can be represented by the formula shown in 1.1. The two horizontal lines represent the bonds of the main chain. The brackets (or parentheses) indicate that the structure repeats many times. The actual number of repeating units present is normally not specified, but is represented by the subscript, n.

Miscellaneous Inorganic Polymers

The polymers discussed in this chapter are simply those that do not fit into one of the earlier, more general categories. They are generally polymers that are not yet of great commercial importance, often because they are so new that much more research needs to be done before they can be utilized effectively. It should be mentioned first that a number of minerals, and glass itself, contain silicon and consist of polymeric structures. Glass is a highly irregular material that consists of rings and linear chains of silicate units in complicated three-dimensional arrangements. On the other hand, some minerals consist of single chains or double chains in which negatively charged oxygen atoms are neutralized by positively charged metal cations. It is sometimes possible to make such materials more tractable by breaking open the cross-links and inserting non-ionic, non-polar groups in their place. Nonetheless, these materials generally are not synthesizable, characterizable, and processable in the way that most organic or inorganic-organic polymers are. The most important class of silicon-containing polymers that have not yet been covered are the polysilazanes, shown. These polymers, or precursors to them, are generally prepared by the reaction of organic-substituted chlorosilanes with ammonia or amines as is shown in reaction.

Polyphosphazenes

Polyphosphazenes comprise by far the largest class of inorganic macromolecules. At least 700 different polymers of this type have been synthesized, with a range of physical and chemical properties that rivals that known hitherto only for synthetic organic macromolecules. Most polyphosphazenes have the general molecular structure. The polymer backbone consists of alternating phosphorus and nitrogen atoms, with two side groups, R, being attached to each phosphorus. The side groups may be organic, organometallic, or inorganic units. Each macromolecule typically contains from 100 to 15,000 or more repeating units linked end to end, which means that (depending on the organic side groups) the highest molecular weights are in the range of 2 million to 10 million. The bonding structure in the backbone is formally represented as a series of alternating single and double bonds. However, this formulation is misleading. Structural measurements suggest that all the bonds along the chain are equal or nearly equal in length, but without the extensive conjugation found in organic polyunsaturated materials. This anomaly will be discussed later. In addition to linear polyphosphazenes with one type of side group, other molecular architectures have also been assembled. These include polyphosphazenes in which two or more different side groups, R1 and R2, are arrayed along the chain in random, regular, or block distributions. Other species exist with short phosphazene branches linked to phosphorus atoms in the main chain. Also available are macromolecules in which carbon or sulfur replace some of the phosphorus atoms in the skeleton. Star-geometry structures, using the symbolism defined, are also accessible. A new and growing area is the field of phosphazene-organic and phosphazene-polysiloxane hybrid linear copolymers, and comb copolymers of the types. In addition, polymers are known in which six-membered phosphazene rings are side groups linked to organic polymers, and where phosphazene rings are linked by organic connector groups to form cyclolinear or cyclomatrix materials.

Characterization of Inorganic Polymers

A wide variety of properties are of interest for the general characterization of polymers, as demonstrated in numerous textbooks and in more specialized books dealing specifically with characterization methods. In addition to the information of this type appearing in this chapter, there is related information in numerous other parts of this book, in particular in Chapters 4 and 8. From any of these sources of information, it becomes immediately obvious that one of the most important properties of a polymer molecule is its molecular weight. This is the characteristic that underlies all the properties that distinguish a polymer from its low-molecular-weight analogues. Thus, one of the most important goals in the preparation of a polymer is to control its molecular weight by a suitable choice of polymerization conditions. Many properties of a polymeric material are improved when the polymer chains are sufficiently long. For example, properties such as the tensile strength of a fiber, the tear strength of a film, or the hardness of a molded object may increase asymptotically with increases in molecular weight, as is shown schematically in Figure 2.1. If the molecular weight is too low, say below a lower limit Ml, then the physical property could be unacceptably low. It might also be unacceptable to let the molecular weight become too high. Above an upper limit Mu, the viscosity of the bulk (undiluted) polymer might be too high for it to be processed easily. Thus, a goal in polymer synthetics is to prepare a polymer so that its molecular weight falls within the “window” demarcated by Ml and Mu. This is frequently accomplished by a choice of reaction time, temperature, nature and amount of catalyst, the nature and amount of solvent, the addition of reactants that can terminate the growth of the polymer chains sooner than would otherwise be the case, addition of complexing agents such as crown ethers, or by the presence of an external physical field, such as ultrasound. The termination of the growth of a particular chain molecule is a statistical process. If termination happens soon after the chain starts to grow then, obviously, the completed chain will be short.